The increasing use of community-based long-term care by a rapidly expanding population of older adults with neuro-psychiatric disorders presents the field of mental health service system research with new challenges. Clinical outcomes and cost-effectiveness research are required to plan adequate combinations of medical, mental health and social welfare services for individuals with varying needs for long-term care. Currently, there is no taxonomy of the patterns or combinations of community-based services which are used by older adults with mental disorders; rather these service patterns are embedded in the broader service categories of """"""""home and community-based services"""""""" or """"""""alternatives to nursing home care."""""""" Using a multi-method approach, this two-year study develops a taxonomy of community-based long-term care services for this older adult population. The significance of the research is its contribution of a tool for service system research necessary to conduct substantive policy studies on the use of community- based services by older persons who have either mental illness or dementia. The overall aim of the research is to construct a valid taxonomy using Colorado's system of long-term care. The proposed study has five steps. The first four steps develop the taxonomy of community- based long-term care and test its validity in local public mental health service areas. The fifth step examines the generalizability of the validated constructed types. These steps: 1) develop a conceptual model of functional types of home and community-based long-term care services; 2) empirically identify patterns of long-term care services used by older adults with neuropsychiatric disorders; 3) examine the congruence of the conceptual and empirical taxonomies utilizing case managers as expert key informants; 4) test the discriminant validity of the typology using person-level characteristics as predictors of service patterns; and 5) assess regional and state-wide generalizability of the constructed types. The study population is Colorado Medicaid clients (60 years and older) with major mental illness or dementia who participate in Home and Community-based Long-term Care Programs; study subjects comprise the client universe on Jan. 1, 1998 who had at least 3 continuous service months. Both qualitative interview data and quantitative client characteristics and service data will be used to develop the taxonomy and test its validity. Secondary data analyses of Medicaid paid claims data will empirically classify patterns of services; multivariate analytic techniques will be employed to test the validity of these classifications. Meta-analytic statistical techniques will be used to examine generalizability of valid classifications. The findings should guide multi-state replications to refine the tool and be used in future research to examine price and functional substitutability of varying community-based models of long-term care that can adequately replace nursing home care for selected types of older persons.
